Contents I. Introduction
High fidelity imaging of moving targets is a well-known challenge for synthetic aperture radar (SAR), especially in ocean environments. Standard SAR processing methods assume the entire scene is stationary, and interpret the Doppler histories in the data accordingly. Any Doppler shifts introduced by target or scene motion will be misinterpreted and the corresponding backscatter will be misplaced in the image. The classic example of this is the ‘train-off-the-track’, in which the signature of a range-traveling train appears displaced from the signature of the track upon which it is running. More serious distortions occur if different parts of the target move at different speeds, which is often the case for vessels moving on a dynamic ocean surface, as the target's signature in the image will then be smeared, not merely displaced. Advanced techniques utilizing multiple-aperture SAR (MSAR) systems have been developed to detect moving targets, estimate their velocities, and correct their positions within the image [1]–[3]. The apertures in these systems are arranged along the flight axis to provide measurements of the scene at slightly different times but from the same vantage point in space, thereby allowing separation of the scene motion from that of the SAR platform. However, all these techniques assume the clutter is stationary, and thus they have limited applicability to ocean environments.
